Julian Davis

Reconstruction of transcriptional dynamics from gene reporter data using differential equations

Motivation: Promoter-driven reporter genes, notably luciferase and green fluorescent protein, provide a tool for the generation of a vast array of time-course data sets from living cells and organisms. The aim of this study is to introduce a modeling framework based on stochastic differential equations (SDEs) and ordinary differential equations (ODEs) that addresses the problem of reconstructing transcription time-course profiles and associated degradation rates. The dynamical model is embedded into a Bayesian framework and inference is performed using Markov chain Monte Carlo algorithms. Results: We present three case studies where the methodology is used to reconstruct unobserved transcription profiles and to estimate associated degradation rates. We discuss advantages and limits of fitting either SDEs ODEs and address the problem of parameter identifiability when model variables are unobserved. We also suggest functional forms, such as on/off switches and stimulus response functions to model transcriptional dynamics and present results of fitting these to experimental data. Contact: b.f.finkenstadt@warwick.ac.uk Supplementary Information: Supplementary data are available at Bioinformatics online.

E18_5 Pituitary Data1

GFP fluorescence measured from single cells in embryonic day18.5 pituitary tissue and subsequent modelling of transcription activity using a stochastic switch model.

Adult Pituitary Trypsin Treated Data2

GFP fluorescence measured from single cells in adult pituitary tissue treated with trypsin for 2hrs prior to confocal imaging. Compare to Adult Pituitary Trypsin Untreated Data2.